Discussions

News: Buzzword watch: Goodbye "Middleware," Hello SOA Applications

  1. Summary
    David Litwack is responsible for the development and advancement of Novell's secure Web services strategy, a position he assumed in July 2002 following Novell's acquisition of SilverStream Software, a company for which he'd served as president and CEO since 1997. He is also a member of Novell's Worldwide Management Committee. JDJ speaks with him about a range of contemporary computing issues.
    Read the interview: Goodbye "Middleware," Hello SOA Applications

    Threaded Messages (29)

  2. I found David's following paragraph especially profound and insightful. He clearly cares about interoperability and mobility. Curiosly vague; no mention of transport, XML, nor Java.
    There were two things wrong with client/server: the client referred to a specific hardware device and the server referred to a specific back-end system. There was a hard wiring between the two. The goal today is to provision any information or systems, regardless of how they are physically implemented, to any audience, regardless of how they connect, in a secure and personalized way based on identity. What we want is to dynamically match the logical service to an identity. In effect, identity/services is a better description of this new architecture.
  3. I am really bored with this SOA hype.

    Yes, SOA is here to stay, but Middleware is not going anywhere. Its going to stay as well.

    My understanding is, SOA is an architecture model where you got to think in terms of Service.
    But to implement your SOA at some point you are going to need middleware.

    WebServices is would be useful in B2B integ, and middleware is going to sit where it is now, A2A integ.

    Am i correct or i am talking something completely rubbish?
  4. It is just a new name for old good things (it depends on SOA definition)
    I am sure we will see new buzzwords to replace WS and SOA too, this always happens.
    BTW, how "Middleware" conflicts with service oriented architectures ?
  5. Middleware is not going anywhere. Its going to stay as well.
    Middleware isn't vanishing. It's becoming more diffuse. It's being componentized and federated.
  6. SOA and Middleware[ Go to top ]

    Hi all

    The title of the article is left of the ball park; SOA and middleware are all of the same thing for many large scale middleware toolsets. A classic example is Quovadx (cloverleaf) for example, where SOA (implemented as webservices) is another wrapper over their already mature integration and BPMS (/workflow) engines. The middleware product provides an essential integration point for large organisations, wrappering this with a webservice layer, and redefining the interfaces to the middleware into something thats a damn site easier to interacte with, aka XML Web Services over HTTP/SOAP, is a natural progression from propritary or less that open middleware integration adapters.

    One important thing to remember about SOA is the guts of what your trying to achieve here. I hear time and time again that webservices will help us solve this and that business or partner integration issue, but the guts of "how", especially in terms of shared security, business process, audit, workflow modules is rarely designed to a point that actually makes investment sence to the underlying business and what we are trying to achieve with SOA. The middleware toolsets really set the stage for this sort of management infrastructure, the presentation layer (aka SOA webservices) is an "cheap" add on.

    Also, dont forget that SOA isnt simply webservices; there are some excellent CEO magazine articles discussing this and its well worth revisting what SOA really is in terms of interface layer to your shared and maaged business processes. All solutions have quite different needs (speed/time/interface/maintenance/etc) and we all need to be open to that fact when building a SOA framework.

    Cheers

    Ck
  7. I do believe that the term SOA is severely over-hyped, especially when it gets constantly linked to Web Services. Web Services as a technology occupies a fairly narrow niche – it gives a standard way to integrate different systems together in a distributed fashion. However, SOA can be easily applied within a system whether it is distributed or not. In a service-oriented application various middleware components are abstracted out as services with clear APIs, lifecycle and configuration management. When used within enterprise (with or without Web Services) SOA allows creating cleanly architected and easily manageable systems.

    I also must disagree with author's analogy stating that middleware market is dying: "People don't buy carburetors to have carburetors. Carburetors are a means to an end. People buy cars." That is just dead wrong. Following this analogy all "Auto-Parts" stores in the world must be on the verge of bankruptcy. The fact is that there is an undeniably huge market for carburetors – they are just bought mostly by car manufacturers.

    To paraphrase the topic of this article, unlike Novell, the middleware market is healthy and booming, and the latest TSS conference can serve as a prove.

    Regards,
    Dmitriy Setrakyan
    xTier™ - Service Oriented Middleware for Java and .NET
  8. The Carburator Shop[ Go to top ]

    "People don't buy carburetors to have carburetors. Carburetors are a means to an end. People buy cars." That is just dead wrong. Following this analogy all "Auto-Parts" stores in the world must be on the verge of bankruptcy. The fact is that there is an undeniably huge market for carburetors – they are just bought mostly by car manufacturers.
    You took his comment out of context and didn't include all of it. The next paragraph is crucial:
    What is the equivalent of the car? An SOA application. At Novell, we've been working to bring together all aspects of identity-based SOA into a suite for that reason. The more transparent we can make middleware, the easier it will be to deliver SOA applications. That will be the new category.
    People don't just go out and buy carburetors, they buy them when they need them in order to make their car work again. it is a secondary market, driven by the primary market which is cars. Same for "middleware", you just don't buy it to have it, you buy it when you need it to make your application work. And the app is SOA.

    -Michael Mattox
  9. The Carburator Shop[ Go to top ]

    You took his comment out of context and didn't include all of it.

    People don't just go out and buy carburetors, they buy them when they need them in order to make their car work again. it is a secondary market, driven by the primary market which is cars. Same for "middleware", you just don't buy it to have it, you buy it when you need it to make your application work. And the app is SOA.
    I don’t really know what you mean by "SOA". If you mean Web Services, then I disagree for the reasons you can find in my original post.

    However, I don’t think I took the comment out of context. The author of the article clearly stated that Middleware market is dying and the only market existing right now is for end solutions. And again I will reiterate that this point is way off.

    We have been hearing stories that Middleware market will go away for ages, but on the contrary, it only grows stronger. Look how many people use Hybernate for example. I would not consider Hybernate a dying product by a far stretch. And yet it is not an end solution – it is a middleware product. The same reasoning can be easily applied to all J2EE Application Servers as well.

    Regards,
    Dmitriy Setrakyan
    xTier™ - Service Oriented Middleware for Java and .NET
  10. OOPs, here comes SOA (again)[ Go to top ]

    Strip all the behavior off of our domain objects. Define data structures to pass around. Provide sets of operations as stateless services, which perform functions on data, and return data to its callers. That sounds revolutionary.

    All we have to do now is eliminate the object-oriented programmers, and the revolution will be complete.
  11. OOPs, here comes SOA (again)[ Go to top ]

    SOA is being hyped so much, that this is the coolest thing happening to s/w sstems design. Hype that atleast the major vendors like IBM , BEA and Msft are trying to create. (With ofcourse lukewarm response from the community!).

    SOA is just a invocation mechanism. Much like any RPC (simple Dec-RPC or the more evolved RMI or CORBA). The hype is like saying 'Iam building a complete banking solution using TCP/IP'. While service orientation is surely an important design consideration, it is not a complete systems-design framework.

    When it comes to new systems (being designed ground up), while SOA has to be factored in into the overall system design, the limitations posed on performance due to distribution will be a limiting factor for Service orientation. Limiting its utility. SOA is good just for interactions of logically 'separate' systems. And not or designing and implementing any and every system!

    Check out more on my blog

    Cheers,
    Ramesh
  12. OOPs, here comes SOA (again)[ Go to top ]

    Alas it will not help. The "Middleware" programmers will always find other areas because they are the equivalent of a hammer searching for nails.

    Look forward to my new play: “How to force object programming into contract consulting “, a comedy in 3 acts.

    Let us all take of our imaginary hats and have one silent minute for all the poor customers.

    BTW, I like your title. Sometimes someone is able to express a whole tragedy into one powerful sentence. But you have distilled it into one word, OOPs.!

    Elegant

    Regards
    Rolf Tollerud
  13. OOPs, it happened AGAIN!!![ Go to top ]

    http://www.smh.com.au/articles/2004/05/25/1085442112928.html
  14. OOPs, here comes SOA (again)[ Go to top ]

    :-)

    To Rolf

    flaf (meaning f***ing laf)
  15. Yo forum..

    I am just tired of this SOA hype, and the various compabies trying to make SOA equal to SOA. It is certainly not. WebServices is a way to get to a SOA - nothing more nothing less.

    It actually reminds me of the way WebServices was "over-hyped" for about two years ago or so...and it still is... Back then, a one of my fellow developers was almost fainting the new faaaantastic thing called WebServices.... I just told him..."It is just another way of exposing your....ahem... scchhh... "functionality" (now people call them services) to the world.." It is just done in an (platform)neutral way. He just wouldn't get it. Being so "brainwashed" by all the hype.

    Back to SOA...Same thing applies.. It has been done for years..some might say decades.. In my opinion SOA is all about architecture... You can get to a SOA by doing a simple J2EE application. Putting the system use case into neat little boxes that applies to a specific problem domain, eg. invoicing, and making a clean interface to that box... Isn't that just what the Facade Pattern does..? YES it is.

    So my main point is:
    SOA is a way of constructing the "infrastructure" of a system, in a maintainable way, with a service layer which all communication goes through.
    It doesn't matter if you use WebServices or call it a SOA or not. It has been done in years, so ...please... stop the hype.

    Am I missing something real bad?.. Maybe I just fell of around the last corner..Dunno.. But please tell me so...if I did

    My pennies

    /Henrik
  16. make SOA equal to SOA.
    That would be "SOA equal to WebServices"
    Sorry.
  17. Henrik: "Am I missing something real bad?.. Maybe I just fell of around the last corner..Dunno.. But please tell me so...if I did"

    I'll be glad to.

    It is not Web Services only that is the revolution, neither RIA (Rich Interface Applications).

    It is the combination of WS and RIA that does the trick. Functional and elegant/beautiful applications that keep the state and leave the server clean and stateless. That is the great paradigm shift. It has already started but will explode when technologies like Avalon/XAML become common.

    WSE 2.0 announced at TechEd yesterday

    Rebecca Dias and Steve Ballmer announced the general availability ("RTM") of the Web Services Enhancements 2.0 toolkit, affectionately known as "wizzy" (or as Steve Ballmer put it, "wussy"). Rebecca demonstrated using WSE 2.0 alongside the Office Information Bridge Framework, making Office more accessible to developers.

    During the announcement, Rebecca demonstrated the use of a Web service from Outlook, using WSE 2.0 to create authorization constraints embedded in a WS-Policy document, and using Outlook's "smart tags" to retrieve interesting financial data to display and interact with as part of the demo


    http://www.technewsworld.com/story/industrynews/33983.html

    MS is going on. The J2EE world is grumbling.

    Regards
    Rolf Tollerud
  18. time to make real applications[ Go to top ]

    Nobody listens to you anymore Rolf. Not that they aren't looking at your posts. But you're not interesting anymore. You said it all. and again
  19. no "I was wrong"[ Go to top ]

    Dorel,

    You know what? You are right.

    Especially that when you are right nobody care, and nobody apologize for blatantly wrong post is the past.

    Regards
    Rolf Tollerud
    (never proved wrong yet)
  20. no "I was wrong"[ Go to top ]

    never proved wrong yet != alway right.
  21. You are missing something[ Go to top ]

    Henrik,

    Don't think of it as the same old thing.

    Think of it as a way to get managers to fund yet another strategic architecture. One where you can lead the team, and break though to bonus by head count.

    It's a great idea, so what if its backed by the slowest protocol in the world, with no established security, or well defined demarcation of internal and external clients/services. These are the holes which allow your vision to add value, and let you lead your team to victory!

    It's all about glass half full/empty. And drinking it as well.

    Jonathan
  22. You are missing something[ Go to top ]

    thanks, that was priceless (if only it wasn't true)

    to follow your lead:

    The vendors think of it as a way to get customers to pay for yet another immature product with a necessary paid upgrade cycle. One where the customer's CTO can add another acronym to their 'career achievements' and the vendor's sales team can break though to bonus via signing count.
  23. You are missing something[ Go to top ]

    Without software fashion and all such 'revolutions' we would still be doing C multithreaded telecoms apps. They never worked anyway. (ahem).

    Change is the one constant, the secret is to enjoy the (b)leading edge. SOA, AOP, AnyOP is good stuff.

    Jonathan
  24. You are missing something[ Go to top ]

    Without software fashion and all such 'revolutions' we would still be doing C multithreaded telecoms apps. They never worked anyway. (ahem)...
    You should rephrase it "Mines never worked anyways". It's not our fault you got it wrong. And there is nothng wrong with C and multithreaded applications. You can anytime switch to LOGO if you like...
  25. Jonathan: "Think of it as a way to get managers to fund yet another strategic architecture."

    Yes, better to go back to EJB Entity beans!
    Read Jonathan Gibbons: Java to J2EE to Oblivion?
    http://www.tallsoftware.com/dotnet/Java%20to%20J2EE%20to%20oblivion.pdf

    And the following TSS discussion:
    http://www.theserverside.com/news/thread.tss?thread_id=14190

    Regards
    Rolf Tollerud
  26. Arrrgggg, yes. Oblivion.[ Go to top ]

    I'm thinking of a follow up to that one. After all its almost 2 years later.

    And the job market is now for C# GUI's talking Soap through to java middle tiers. That paper was still the most fun I've had here, and I learnt loads from the better contributers about the role of Linux in corporate strategies. Or rather the potential. And the hype for the last year has all been about simple frameworks (spring et al), and open/simple OR layers (hybernate, emeraldjb..)

    BTW, no matter what you think of Rolfs comments, its folk like him who stick with the community (for whatever reason) that make it a community. Staying power counts for a lot.

    Jonathan
  27. Arrrgggg, yes. Oblivion.[ Go to top ]

    ...And the job market is now for C# GUI's talking Soap through to java middle tiers...
    and 10-12 years ago, it was the same. Diferent products and brands but still the same: VB talking to MQSeries, PowerBuilder talking to CICS, PowerBuilder talking to CICS, VB talking to CICS thru SNA, MicroFocus Cobol talking to mainframes/devices using CORBA... and now C# GUI talking to Java thru SOAP.

    Is that really innovation?
  28. Arrrgggg, yes. Oblivion.[ Go to top ]

    Yep. Thats innovation. A congealing of technologies over time. Allowing slicker stuff to come out and enter the developer world.

    What about Lamp eh? That's fear of god stuff. Just thank god they don't have a gui language (go on, someone correct me).

    Jonathan
  29. Arrrgggg, yes. Oblivion.[ Go to top ]

    ...BTW, no matter what you think of Rolfs comments, its folk like him who stick with the community (for whatever reason) that make it a community....
    Even good families have their own black sheep...
  30. SOA Applications ?[ Go to top ]

    Rich clients invoking well defined services... It reminds me of my first job ten years ago or so... VB clients calling a neat Tuxedo backend.

    The point of SOA and Webserver is not for buildind applications but for the interoperation of lowly coupled applications. In intra-application usage there is no need for an standard XML/SOAP. There is a huge need for performance and transactional state.

    SOA as a way of architecting applications... well, the same old theme.... Tuxedo services, service beans, etc.... just put a Webservice layer in front, and you can sell the last decade most radical architecture hype.

    The identity interchange/management of application interoperation is the really interesting point.

    Regards.

    Josep Blazquez